Accumulating epidemiologic and experimental evidence suggests a possible association between diabetes mellitus and infection with common viruses, particularly members of the Coxackievirus and Myxovirus groups. This epidemiologic investigation is based on the hypothesis that naturally-occurring virus infections cause damage to the beta cells of the islets of Langerhans and juvenile-type diabetes mellitus in man. In this project, the prevalence of serum antibodies against selected viruses will be determined in persons with recent-onset diabetes mellitus and appropriately matched controls. In addition, tissue typing for HL-A antigens and tests for anti-tissue antigens will be done. Diabetics between the ages of 5 and 21 years will comprise the study group; control subjects will be selected from three groups; a) siblings, b) patients of the referring physician and c) close associates of the index case. Serological studies will be carried out by standard techniques using complement fixation and neutralization tests. The study is designed to determine by appropriate statistical evaluation whether or not patients with juvenile-type diabetes experience more infections, or more severe infections, with common viruses than do normal individuals.